Schema inference

Vadalog language supports data source binding through the @bind annotation. Vadacode improves user experience with binding with schema inference, i.e. the ability to automatically infer the schema of the data source and enrich with it the binding.

Schema inference is available for CSV and Apache Parquet input data sources.

Infer schema from input data

Vadacode adds two contextual menus to the VSCode Explorer: when you right click on a CSV or a Parquet file, you are prompted with a Vadacode: add to CSV/Parquet action. Just click on it, and Vadacode will create a new binding at the current cursor position in the active Vadalog file. With a single click, you will get:

  • the @input statement for a new predicate after the name of the file;

  • the @bind annotation already configured with the relevat paths;

  • the @mapping statements for each column in the file, with the column name and type inferred from the file;

  • the documentation for the predicate, with the column names and types, in Vadoc format.

schema inference
Figure 1. Vadacode schema inference